The farming societies that developed in southern Europe, particularly in the Iberian Peninsula, from the mid-sixth millennium, began to experience a huge social expansion (including demographic growth, intensification of the agricultural economy, copper metallurgy, supra-regional exchange networks, etc.) by the last quarter of the fourth millennium (García Sanjuán and Murillo-Barroso, 2013). One particular indication of these changes during the Copper Age was the acquisition of a great number of exogenous raw materials and the appearance in funerary contexts of “special” individuals or collectives, buried with sophisticated grave goods. This phenomenon has been interpreted as indicative of an emerging “elite”, something that is quite palpable in the lower Guadalquivir River valley during the Chalcolithic Period (between 3200 and 2200/2300 cal BC). The presence of ivory objects seems to be one of the common denominators in local elite grave goods there (García Sanjuán et al., 2018b; Luciañez-Triviño et al., 2021). The presence of significant amounts of ivory, and other extra-peninsular materials (such as ostrich eggshell) in some of the largest Chalcolithic settlements in Iberia reflect the intensification of these trans-Mediterranean exchange networks.

Ivory has been a widely appreciated raw material since the Paleolithic and throughout human history. In the Iberian Copper Age, elephant ivory was used to produce many different kinds of prestigious objects such as handles, zoomorphic and anthropomorphic figurines, combs, decorated and undecorated plates, vessels, boxes, lunulae, and beads. Ivory was not used during the Chalcolithic to make utilitarian objects, like tools to produce other goods, but exclusively for showy personal objects for display, as body ornaments or to be carried by individuals (Luciañez-Triviño et al., 2021).

The study of this raw material is at the core of current research on resources in Iberian prehistory, providing as it does, valuable information about social change, crafts, long-distance contacts, and socio-cultural organisation. Since the end of the nineteenth-century scholars have noted the presence of ivory on Iberian sites from the Late Neolithic to the Iron Age. The study of this raw material still raises a great deal of interest even today, something reflected in a vast literature that greatly exceeds the scope of this paper. Despite the extensive literature about these Iberian ivories, there has been less research on ivory manufacturing and technology (e. g. Altamirano García, 2014a, 2014b) ¹ . Most references deal with material from later periods, mainly the Bronze and Iron Ages (e. g. Barciela González, 2006, 2007, 2012 ² ; Altamirano García, 2011, 2012a, 2012b; López Padilla and Hernández Pérez, 2011; Liesau von Lettow-Vorbeck and Schuhmacher, 2012; López Padilla, 2012; Pascual Benito, 2012; Blasco Martín, 2015; Pau et al., 2018). In contrast, the study of ivory technology elsewhere in Europe is most developed for the Palaeolithic period (e. g. Hahn et al., 1995; Christensen, 1999; Khlopatchev, 2001; Heckel, 2009; Khlopatchev and Girya, 2010; Hein, 2011, 2014), with contributions including the experimental approach (e. g. Hahn, 1986; Hahn et al., 1995; Christensen, 1999; Malina and Ehmann, 2009; Khlopatchev and Girya, 2010; Heckel and Wolf, 2014).

This article explores the evidence for ivory technology during the Chalcolithic, based on the materials from the site of Valencina de la Concepción-Castilleja de Guzmán (hereafter Valencina) in southern Spain. The results concerning production modes, techniques, and the tools used in manufacturing these objects will be discussed in this article, in order to understand all aspects involved in the production of ivory artefacts, from the supply of the raw material, manufacturing processes, to their end use (the object biography), with special attention to debitage (blanks production). Thus, all products from the technical chain ‒ blocks, blanks, pre-forms (or roughouts), and production waste (or debris) ‒ were studied.

Valencina is located in North Aljarafe, ca. 8 km from the city centre of Seville (Fig. 1). The radiocarbon dates place the activity there between ca. 3200 cal BC and ca. 2300 cal BC (García Sanjuán et al., 2018a). This site is considered to have played a prominent role in the networks at the time due to its proximity to multiple resources of economic and social value as well as the abundant flow of exogenous materials (e. g., ivory, amber, ostrich eggshell, etc.) (Vargas Jimenez, 2003; Vargas Jimenez et al., 2010; García Sanjuán, 2017). The intense research carried out at Valencina over the last decade has brought advances in human anthropology (e. g. Lacalle Rodríguez et al., 2000; Robles Carrasco and Díaz-Zorita Bonilla, 2013; Pecero Espín, 2016; Díaz-Zorita Bonilla, 2017; Robles Carrasco et al., 2017; Cintas Peña et al., 2018) as well as in faunal studies (e. g. Liesau von Lettow-Vorbeck et al., 2014) ³ , chronology (García Sanjuán et al., 2018a) and its material culture (e. g., Hunt Ortiz and Hurtado Pérez, 2010; Hunt Ortiz et al., 2011; Rogerio-Candelera et al., 2013; Odriozola Lloret and García Sanjuán, 2013; Schuhmacher et al., 2013a; Luciañez-Triviño et al., 2014; García Sanjuán et al., 2016; Luciañez-Triviño and García Sanjuán, 2016; Morgado Rodríguez et al., 2016; Murillo-Barroso, 2016a, 2016b). For a historiographic synthesis see e. g. García Sanjuán (2013) and García Sanjuán et al. (2018a). Some papers regarding typology, contexts, or raw material characterization of Valencina’s ivories have been published (e. g. Schuhmacher, 2012; Vargas Jiménez et al., 2012; García Sanjuán et al., 2013; Nocete Calvo et al., 2013; Schuhmacher et al., 2013a; Luciañez-Triviño et al., 2014; Luciañez-Triviño and García Sanjuán, 2016; García Sanjuán et al., 2018a). Very little, however, is yet known about technology’s specifics.

A total of 8.8 kg of ivory has been found at Valencina. To date, it represents the largest published ivory assemblage in the Iberian Peninsula and Europe for this period. The ivories were recovered from 12 Chalcolithic structures coming from eight sectors of the site (for a description of the archaeological contexts see Luciañez-Triviño et al., 2021). Ivory accounts for 65% of the whole osseous industry in the structures where ivory is present, followed by objects made from animal bone (21%). With only a few exceptions, the ivory generally appears in large burial monuments (mainly tholoi).

The ivory material was found in a very poor state of preservation so the assemblage was highly fragmented. This circumstance imposes limits on the study, making it difficult to get a true picture of the number of analytical categories that actually would have been present. However, previous restoration work on several of the artefacts permitted a more realistic picture and also allowed observation of the technological traces hidden by dirt, sediment, etc.

In order to identify the raw material, the entire ivory assemblage was carefully examined with a stereomicroscope Nikon SMZ800 (up to 126X), and a digital microscope ShuttlePix P-400R (up to 400X). In order to discriminate between raw materials and to identify the artefacts made from ivory, these observations were compared with our own reference collection and with the structural characteristics of different raw materials based on what is available in the scholarly literature (e. g. MacGregor, 1985; Krzyszkowska, 1990; Espinoza and Mann, 1991, 1993, 1999; Haynes, 1991; Deschler-Erb, 1998; Christensen, 1999; Feldhamer et al., 1999; Kardong, 1999; Ábelová, 2008; Locke, 2008; Rijkelijkhuizen, 2008; Virág, 2012; Choyke and O’Connor, 2013; Tolksdorf et al., 2015). Additionally, for a better understanding of the observed tool traces, some focussed experiments were carried out. The specific characteristic of the proboscidean ivory and its cracking patterns were studied in depth in order to infer the relative position of the object within the tusk as well as the production modes.

The material was catalogued according to five categories of analysis: blocks, blanks, pre-forms/rough-outs, debris/production waste, and finished objects/end products. The definition of the categories of analysis (Fig. 2) are as follows (some definitions and categories were adapted from the work of A. Averbouh, 2001): a primary block is the complete anatomical support, that is, the complete tusk. A secondary block (a, more or less, large segment of the initial block) or a prepared block (a primary block where the unwanted or useless parts have been removed) can be extracted from the primary block. Blanks/supports are elements displaying little manipulation or transformation. They derive from the debitage and are obtained from primary or secondary blocks. These items are produced with the direct and specific intention of manufacturing a particular object. The pre-forms/rough-outs occupy an intermediate position between the support and the finished object. They have been more intensively manipulated, and already contain an approximation of the morphology of the final object. The debris/production waste comes from all the stages in the ivory working. Their production was not intentional but resulted from the transformation operations and includes chips, flakes, and fragments. The description of the surface technical traces and general vocabulary for technology is based on Averbouh and Provenzano (1998-1999) and the Multilingual Lexicon of Bone Industries (Averbouh, 2010). Cataloguing and definitions of finished objects and decorations are based on Luciañez-Triviño ⁴ and Luciañez-Triviño et al. (2021).

A total of 242 items were studied. Within the assemblage, there are Blocks (N = 6), Pre-forms/Rough-outs (N = 1), Debris/production wastes (N = 14), and Finished objects/End products (N = 175) (Table 1). Possible unintentional errors or omissions during excavation must be considered when identifying and recovering small fragments, such as chips. Besides, the excavation report is not specific about the structures whose contents were sieved ⁵ . Consequently, these numbers may not correspond to the totality of the material culture that may have been present. There are a huge variety of finished objects such as Receivers (rectangular boxes, containers with cylindrical mouths, curved/oval based vessels, cornucopia-like objects, handles, hilts, and “others”); Perforated objects (rings, bracelets, perforated discs, and beads); Toothed elements (combs and ornamental combs); Figures (human, animal, vegetal figurines or other shapes) and Plates (see Luciañez-Triviño et al., 2021).

In order to understand the methods of exploitation, it was necessary to develop an in-depth knowledge of the characteristics of the ivory that could be observed on the artefacts. Therefore, part of the research focused on identifying the features of the ivory that can be observed by the naked eye or at low magnifications. The most relevant characteristics observable in each of the cuts or views are described in the following paragraphs.

Some vertebrate teeth have been enlarged or lengthened in response to very different evolutionary processes into what are commonly called “tusks.” Such is the case of extant proboscideans and their extinct relatives. Elephant tusks are the second continuously growing maxillary or upper incisors (I2) (Feldhamer et al., 1999, p. 314; Virág, 2012, p. 1406). Proboscidean ivory has a complex architecture characterized by successive steps or stages of organization (Locke, 2008). This issue has been addressed by multiple scholars resulting in a variety of proposed growth models (e. g., Miles and White, 1960; Raubenheimer et al., 1990; Raubenheimer et al., 1998; Espinoza and Mann, 1999; Su and Cui, 1999; Palombo and Villa, 2001; Nalla et al., 2003; Trapani and Fisher, 2003; Ábelová, 2008; Locke, 2008; Virág, 2012; Albéric, 2014).

The physical structure of the elephant tusk, described from the inside out comprises: the pulp cavity, dentin, cementum and enamel (for a better understanding of the anatomy of this tooth consult: Virág, 2012, fig. 1). The cross-section of the tusk is round or oval. For the most part, it is composed of dentin, covered by a layer of cementum, which can be very thick and have a laminar appearance in the case of extinct genera (Espinoza and Mann, 1991). At the tip, there is a conical enamel layer, which, however, is worn away during the life of the animal (Haynes, 1991). Consequently, it is not usually found in archaeological material, let alone in very elaborate artefacts. Elephant ivory has an advantage over other raw materials because of the amount of material it provides (dentine constitutes 95% of the tusk), especially in tusks from older animals. The proboscidean tusk is a continuously growing tooth, so the older the animal, the more raw material the tusk provides. Dentine is a mineralized connective tissue with an organic matrix of collagen proteins (Roylance, 2000-2001). It is a very durable material because the collagen fibres reinforce the inorganic matrix and give it elasticity (Virág, 2012, p. 1406), while the inorganic part gives ivory strength and rigidity, and its laminar configuration and tubule arrangement make it resistant to flexion (Locke, 2008, p. 447).

Here, four cuts or views of ivory with specific characteristics have been identified to aid in a first taxonomic identification and for the technological study: 1) Transverse plane/cross-section; 2) Natural tangential plane; 3) Artificial tangential plane; and 4) Radial plane.

The characteristics set out in this paper are observable both in fresh or well-preserved ivory, as well as in archaeological or deteriorated ivory. In the case of the latter, as we outlined, these pieces develop crack patterns that reveal the underlying, specific structure of the ivory as raw material. Knowledge and recognition of these characteristics and structures are essential in order to recognise the ways ivory was worked and how the blanks were extracted from the tusk. Observing these details in archaeological artefacts and particularly in the finished objects themselves, helps to identify the position, at least the relative position, of the blank (and the object) and where it was taken out from the tusk. In short, both the characteristics described above and the proposed methodology and vocabulary can be applied to any material, regardless of the period. The use of this methodology could facilitate knowledge of ivory craftsmanship generally and allow assemblages to be compared.

The traces still preserved on the artefacts and the production waste enabled the reconstruction of part of the sequence of ivory working in the third millennium BC in the Iberian Peninsula. The tool kit used by the craftspeople was much more extensive than has been generally thought. The ivory processing involved metal and lithic tools, various abrasives, cords (possibly made of plant matter), as well as probably many other supporting materials made of wood, such as workbenches, grips, etc. The use of abrasive sawing dates to approx. 4000-3200 BC, as the evidence described indicates, including this study. The objects with abrasive sawing marks on them come from structure 10.042-10.049, whose first use dates back to around 3725 cal BC (García Sanjuán et al., 2018a). The use of this technique could be traced back to the inefficiency of copper-based saws from the Chalcolithic period, at least as far as the material culture of the Iberian Peninsula is concerned. This inefficiency would have led to the search for a more efficient method of working on osseous materials. The author demonstrated experimentally that the model based on the Valencina-type saws that were used was not able to penetrate the ivory. On the contrary, experimentation with abrasives and cords proved to be a very effective segmentation method ⁶ .

All parts of the tusk can be exploited in such a way as to maximise its productivity and therefore the final object to be manufactured determined the part of the tusk that would be exploited. It can be concluded that there was standardised exploitation during the Chalcolithic period. The standardization was aimed at the manufacture of particular blanks, based on the knowledge and use of the different tusk parts (hollow and compact) in a differential and specific way. Thus, the proximal part of the tusk was specifically used for the manufacture of objects that required a hollow internal area while the distal area of the tusk toward the tip was used for the manufacture of solid and round objects and plates.

The transversal mode was the first kind of exploitation with the aim of segmenting the tusks and obtaining, more or less, thick, compact, or hollow portions. Sometimes segmentation was carried out to facilitate the handling of the tusk (which can be very heavy and large) and in other cases to obtain secondary blocks to produce supports from which vessels, cylindrical boxes, or rectangular boxes were made.

The longitudinal method of exploitation is clearly independent of the first transverse mode, even though it is necessary for the first instance to obtain segments of ivory by segmentation. Conceptually, what was sought after were flat rectangular supports to manufacture flat objects such as combs or plates, but also all kinds of beads or discs.

Among the pieces studied here, 52.5% were manufactured on longitudinal blanks. This proportion makes it appear that both modes of exploitation were used to the same extent, but this is an unrealistic picture when all the indeterminate pieces are counted (N = 67). If only those items whose mode has been clearly identified are considered (Transversal, N = 48; Longitudinal, N = 127; Total, N = 175) then a truer picture emerges: Altogether, 72.6% of the objects are manufactured on longitudinal plates, while 27.4% are produced from transversal blanks.

The data obtained clearly show that longitudinal mode was the preferred method, probably because it permitted the greatest number of objects to be obtained, even in the production of disc-shaped objects for which it would seem to be more convenient and obvious to use thin slices. The discs with central perforations found in Montelirio tholos were made this way (Fig. 9). They were manufactured from longitudinal plates and clearly not from slices. The discs appeared in the Large Chamber of Montelirio in even numbers (N=10) and are “paired” according to their size (see Table 3): four larger discs with very regular diameters of between 40 and 41.25 mm (in decreasing order of size: CG/328-61, CG/328-60, CG/328-60 and CG/327-1): CG/328-61, CG/328-60, MONT-DJ09/19-CG/337-1 and CG/328-138); two medium disks of ø32.8 and ø33.45mm (CG/337-2 and CG/337-3) and four small ones of diameters between 22.26 and 23.3 mm (in decreasing order: CG/318-104, CG/328-105, CG/318-102 and CG/318-103). Not all discs display traces of cementum, so they were not all produced from precisely the same area of the tusk. The study of the disc sizes, the absence or presence of cementum, its location, and extent permit reconstruction of the manufacturing sequence and propose with some reliability that, in this case, all 10 objects may have been efficiently extracted from the same segment of a tusk, that is, from a cylindrical blank, as an example of the high performance of the procedures and methods employed (Fig. 10).

This manner of working the material is several millennia ahead of what in prolific areas such as Anatolia would not be seen until the end of the Bronze Age. A similar level of exploitation can only be found in the eastern Mediterranean and from much more recent periods. For example, a set of decorative plaques from Tiglat-pileser II-I’s palace (745-727 BC) at Ugarit, the modern city of Arslantaş (De Pietri, 2020) revealed a similar kind of production method. The plates, used to decorate some kind of furniture, possibly a bed, were extracted longitudinally, possibly from a single large tusk, or more likely, from a pair of tusks (Caubet and Poplin, 1987, fig. 16, p. 287). Nevertheless, similarly but in a different way, high-performance exploitation was also characteristic of hippopotamus ivory in the same area. In Syria and Anatolia, hippopotamus ivory, rather than elephant ivory, was the preferred raw material from the IV millennium BC and until the end of the Bronze Age, as evidenced by some objects coming from IV and III millennium, high-status contexts (Schwartz et al., 2003; Frangipane, 2004; Caubet, 2013, p. 450; De Pietri, 2020). Later, from contexts dating to around 1800 BC, craftspeople cleverly exploited the anatomic possibilities of the hippo tusks: the proximal area to produce plaques and the distal (solid) region for creating objects with three-dimensional volume. However, two major technical changes took place in ivory working during the Late Bronze period: 1) the number of artifacts made from ivory increased and 2) the use of elephant ivory was introduced on a large scale, although without replacing hippo ivory (Caubet, 2013, p. 451). In the Iberian Peninsula, however, from the III millennium BC, elephant ivory was the dominant material, with very occasional exploitation of other types of ivory (e. g., sperm whale; Schuhmacher et al., 2013b).

On the other hand, the question of the local or non-local production of objects is a recurrent debate among scholars. The main and general problem lies in linking (that is, whether the pieces belong to the same chaîne opératoire) the wastes in so-called workshops with the objects in the tombs, something which is often not possible, or has simply not been attempted. In the case in point, there are some indications that this ivory material was worked locally but to what extent is yet unknown. In any case, local working of ivory could not have started before ±2800 cal BC. Two radiocarbon dates are available directly on two pieces of ivory waste from Structure 402 (IES Sector). Both fragments could derive from the same episode of ivory working dating back to 2855-2580 cal BC (2σ) (García Sanjuán et al., 2018a). The activity in the tholos of Montelirio, dates to 2875-2635 cal BC (2σ), while the use of Structure 10.042 (the first section of structure 10.042-10.049) yielded an older dating of 3725-1840 cal BC (2σ) (García Sanjuán et al., 2018a). The recurrent failure to date some sectors and structures containing ivory objects in Valencina (García Sanjuán et al., 2013, 2018a) makes it difficult to get a clearer vision of the arrival, use, and consumption of ivory as well as the production of objects. Given the current data, a much more varied and fluid scenario should be considered. Before the 29^th century BC this raw material, together with several finished or semi-finished objects and the know-how of how to manufacture it, could have come to Valencina from other areas in the Mediterranean. Subsequently, local craftspeople, or even foreign craftspeople established in the area, could have worked the imported ivory locally.

Much still needs to be learned about ivory craftsmanship in general and in particular around the Mediterranean because of a lack of technological studies. Missing information makes it impossible or very difficult to compare assemblages and therefore to observe the emergence, evolution, or change and the transfer of knowledge of this craft. Nevertheless, it can be said that from its beginnings, ivory working emerged as a specialised craft, with highly refined procedures, possibly based on previous manufacturing experience with working other materials such as wood and bone.